Image Compression-Rangkuman Eko Zulkaryanto PDF
Image Compression-Rangkuman Eko Zulkaryanto PDF
(Bp Aziz)
Dirangkum oleh: Eko Zulkaryanto
(http://zulkaryanto.wordpress.com)
Computer Science Bogor Agricultural University (http://www.ipb.ac.id)
Houffman Coding
That is, 20% of pixels in the image have grey value 50; 40%
have grey value 100, and so on.
Coding redudancy
Interpixel redudancy
Psychovisual redudancy
Kompresi data merupakan usaha untuk mengurangi satu atau
beberapa redudansi tersebut
1.1. coding redudancy
dengan L
: banyaknya gray level
nk
: banyaknya gray level ke-k
n
: banyaknya piksel total
Notice that the longest code words are associated with the
lowest probabilities.
Dengan A(n)
Interpixel redudancy disebut juga spatial redudancy
Salah satu metode kompresi citra untuk menghilangkan
interpixel redudancy adalah run-length coding
Misalkan sebuah citra dengan gray level berikut
4444433333377722222
222222211111 1166666
...
Maka run-length coding dari baris ke-1
(4,5)(3,6)(7,3)(2,5)
Maka run-length coding dari baris ke-2
(2,7)(1,7)(6,5)
Run length encoding
(n)
A(n)
A(0)
A(n)
1
N n
N 1 n
y 0
f ( x, y) f ( x, y n)
KOMPRESI JPEG
00010
00000
Dimana
LL : Low Low Frequency (most importance)
HL : High Low Frequency (lesser importance)
LH : Low High Frequency (more lesser importance)
HH : High High Frequency (most less importance)
luminance
RGB Components
Y U V (Y Cb Cr) Components
AC Coefficient Coding
the 63 quantized coefficients are converted into a onedimensional zig-zag sequence to increase the run length of zeros.
- zero run-length encoding is then applied.
- Huffman coding or arithmetic encoding is then further applied.
and then taking the DCT and rounding to the nearest integer
results in
Maps 8 x 8 to a 1 x 64 vector.
Note the rather large value of the top-left corner. This is the DC
coefficient. The remaining 63 coefficients are called the AC
coefficients.
Step 3: Quantization
The human eye is good at seeing small differences in
brightness over a relatively large area, but not so good at
distinguishing the exact strength of a high frequency
brightness variation.
This fact allows one to get away with greatly reducing the
amount of information in the high frequency components.
This is done by simply dividing each component in the
frequency domain by a constant for that component, and then
rounding to the nearest integer.
This is the main lossy operation in the whole process.
As a result of this, it is typically the case that many of the
higher frequency components are rounded to zero, and many
of the rest become small positive or negative numbers.
A common quantization matrix is:
26,
3, 0,
3, 2, 6,
2, 4, 1, 4,
1, 1, 5, 1, 2,
1, 1, 1, 2, 0, 0,
0, 0, 0, 1, 1, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0,
0, 0, 0, 0,
0, 0, 0,
0, 0,
0
DC coefficient is -26
Zero run-length encoding is
(0, -3),
(1, -3), (0, -2), (0, -6), (0, 2), (0, -4),
(0, 1), (0, -4), (0, 1), (0, 1), (0, 5), (0, 1),
(0, 2), (0, -1), (0, 1), (0, -1), (0, 2),
(5, -1), (0, -1),
(0, 0)
Decoding
a)
b)
c)
d)
e)
JPEG 2000
JPEG 2000 merupakan pengembangan kompresi JPEG.
Didesain untuk internet, scanning, foto digital, remote sensing,
medical imegrey, perpustakaan digital dan e-commerce
Kelebihan :
Dapat digunakan pada bit-rate rendah sehingga dapat digunakan
untuk network image dan remote sensing
Menggunakan Lossy dan loseless tergantung kebutuhan
bandwidth. Loseless digunakan untuk medical image
Transmisi progresif dan akurasi & resolusi pixel tinggi
Menggunakan Region of Interest (ROI)
Robustness to bit error yang digunakan untuk komunikasi
jaringan dan wireless
Open architecture : single compression/decompression
Mendukung protective image security : watermarking, labeling,
stamping, dan encryption
Mendukung image ukuran besar 64k x 64k, size up to 232 - 1
Mendukung meta data dan baik untuk computer-generated
imagenary. Dulu JPEG standar baik untuk natural imagenary.
2-D Wavelet Transform via Separable Filters
Original
Latihan